How to Calculate Vig: Formula + Code

Vig is the tax on every bet you place. Calculating it takes 30 seconds and tells you exactly how much the sportsbook is charging you. This guide covers the formula, code, and the breakpoints that separate sharp books from overpriced ones.

The Vig Formula

Vig — short for vigorish, also called juice, overround, or the bookie’s margin depending on where you are in the world — is the sportsbook’s built-in edge on every market. For the full story of how a Russian word for “winnings” became American bookmaker slang, see The Vigorish: A Linguistic Journey.

Vig calculation works identically across all odds formats. The core idea: a fair market has outcomes summing to exactly 100% probability. Sportsbooks inflate that total. The amount above 100% is the vig.

Step 1: Convert Odds to Implied Probability

American odds (negative):

Implied Probability = |odds| / (|odds| + 100)

Example: -110 → 110 / 210 = 0.5238 = 52.38%

American odds (positive):

Implied Probability = 100 / (odds + 100)

Example: +120 → 100 / 220 = 0.4545 = 45.45%

Decimal odds:

Implied Probability = 1 / decimal_odds

Example: 1.91 → 1 / 1.91 = 0.5236 = 52.36%

Step 2: Sum All Probabilities

For a two-way market (spread, total, moneyline with two outcomes):

Total = P(Side A) + P(Side B)

Example: -110 / -110 → 52.38% + 52.38% = 104.76%

For a three-way market (soccer moneyline with draw):

Total = P(Home) + P(Draw) + P(Away)

Step 3: Vig = Total - 100%

Vig = 104.76% - 100% = 4.76%

That’s it. The book is charging 4.76% on this market.

Worked Examples

Standard Line: -110 / -110

Side A: 110 / 210 = 52.38%
Side B: 110 / 210 = 52.38%
Total: 104.76%
Vig: 4.76%

This is the default price at most regulated US sportsbooks for spread and total bets.

Reduced Juice: -105 / -105

Side A: 105 / 205 = 51.22%
Side B: 105 / 205 = 51.22%
Total: 102.44%
Vig: 2.44%

BetAnything (formerly BetAnySports) and LowVig.ag offer this on standard sides and totals. Half the vig of mainstream books. For a full breakdown, see our guide to reduced juice sportsbooks.

Uneven Moneyline: -140 / +120

Favorite: 140 / 240 = 58.33%
Underdog: 100 / 220 = 45.45%
Total: 103.79%
Vig: 3.79%

Heavy Favorite: -300 / +240

Favorite: 300 / 400 = 75.00%
Underdog: 100 / 340 = 29.41%
Total: 104.41%
Vig: 4.41%

High-Vig Prop: -130 / +100

Favorite: 130 / 230 = 56.52%
Underdog: 100 / 200 = 50.00%
Total: 106.52%
Vig: 6.52%

Player props regularly carry 6-12% vig. This is where books make the most margin.

Calculating No-Vig Fair Odds

To find what the odds should be without the sportsbook’s margin, divide each probability by the total.

For -140 / +120 (total 103.79%):

Fair probability A: 58.33% / 103.79% = 56.20%
Fair probability B: 45.45% / 103.79% = 43.80%
Fair odds A: -(56.20 / (100 - 56.20)) × 100 = -128.3
Fair odds B: ((100 - 43.80) / 43.80) × 100 = +128.3

The fair line is roughly -128 / +128. The book moved it to -140 / +120 to extract 3.79% vig.

Python Code for Betting Agents

AI betting agents calculate vig programmatically across every book before placing a bet. Here’s the core function:

def calculate_vig(odds_a: int, odds_b: int) -> dict:
    """Calculate vig from American odds for a two-way market."""

    def implied_prob(odds: int) -> float:
        if odds < 0:
            return abs(odds) / (abs(odds) + 100)
        else:
            return 100 / (odds + 100)

    prob_a = implied_prob(odds_a)
    prob_b = implied_prob(odds_b)
    total = prob_a + prob_b
    vig = (total - 1) * 100

    # No-vig fair probabilities
    fair_a = prob_a / total
    fair_b = prob_b / total

    return {
        "vig_pct": round(vig, 2),
        "implied_a": round(prob_a * 100, 2),
        "implied_b": round(prob_b * 100, 2),
        "fair_prob_a": round(fair_a * 100, 2),
        "fair_prob_b": round(fair_b * 100, 2),
        "total_implied": round(total * 100, 2),
    }

# Examples
print(calculate_vig(-110, -110))
# {'vig_pct': 4.76, 'implied_a': 52.38, 'implied_b': 52.38, ...}

print(calculate_vig(-105, -105))
# {'vig_pct': 2.44, 'implied_a': 51.22, 'implied_b': 51.22, ...}

print(calculate_vig(-140, 120))
# {'vig_pct': 3.79, 'implied_a': 58.33, 'implied_b': 45.45, ...}

An agent’s line-shopping loop looks like this:

from odds_api import get_odds  # The Odds API client

def find_best_vig(event_id: str, market: str) -> dict:
    """Find the sportsbook with lowest vig for a given market."""
    all_odds = get_odds(event_id, market)
    best = None
    for book, odds in all_odds.items():
        result = calculate_vig(odds["home"], odds["away"])
        if best is None or result["vig_pct"] < best["vig_pct"]:
            best = {**result, "book": book}
    return best

For the full API integration, see our Prediction Market API Reference.

Vig and Break-Even Win Rates

Vig directly determines how often you need to win to break even.

Some books quote vig as a share of total implied probability (hold), which reads slightly lower — 4.55% instead of 4.76% at -110. This guide uses the overround convention throughout.

The difference between -105 and -110 looks small — 1.16% on the break-even rate. It isn’t. At $100 per bet, a 50/50 bettor expects to lose $4.55 per bet at -110 and $2.38 at -105. Over 1,000 bets that gap is roughly $2,160 — pure savings for placing the identical wagers at a reduced-juice book.

MLB is the cheapest major US sport to bet: moneylines at sharp books commonly run around 2.5% vig thanks to dime lines (10-cent spreads between favorite and underdog), the lowest of the major US sports.

Vig Reference Tiers

Based on the AgentBet Vig Index, sportsbooks fall into clear tiers. For a sport-by-sport breakdown, see best sportsbook odds by sport.

Prediction markets charge their “vig” differently: on Kalshi and Polymarket it’s the bid-ask spread plus taker fees — Kalshi charges 0.07 × P × (1−P) per contract and Polymarket US charges 0.06 × P × (1−P). See Prediction Market Odds Explained for how those fees convert to an effective vig.

What’s Next

• AgentBet Vig Index — Monthly vig tracking across 16 sportsbooks
• Prediction Market Odds Explained — Converting between probability, decimal, and American odds
• Sharp Betting Strategies — Exploiting vig differentials for profit
• Agent Betting Glossary — Complete terminology reference
• Agent Betting Stack — How vig fits into the full agent architecture
• Juice Comparison Across Offshore Books — Programmatic vig analysis with Python
• Vig Shopping Strategy — How to exploit vig differentials for profit
• Odds Boost Math — Apply vig math to evaluate sportsbook odds boosts